CV
Download PDF → · Last updated May 2026
Experience
Senior Data Scientist — TapNation
Mobile Gaming · 2022 – Present · Remote/Rome
- Built a cohort-based LTV prediction system using dual ML approaches — ARPDAU × Retention with power-law curve fitting alongside coefficient-based normalized revenue models — with Monte Carlo sampling for confidence intervals, deployed on AWS Lambda for 100+ partner studios.
- Built revenue and margin projection dashboards (Streamlit) for scenario-based forecasting across D1-to-M11 user lifecycle, supporting a portfolio surpassing 1B downloads.
- Designed a real-time bid and ROAS optimization engine serving 10+ ad networks (Meta, TikTok, Google, Unity, AppLovin), with adaptive step-sizing for CPI / t-ROAS / CPE campaigns — deployed as AWS Lambda functions.
- Built Marketing Mix Modeling pipelines (LightweightMMM, Robyn) to quantify organic vs. paid revenue and optimize spend allocation across channels.
- Developed an LLM-powered Text2SQL system enabling non-technical teams to query business data via natural language, materially reducing ad-hoc analytics load on the data team.
- Led the user-acquisition analytics that supported TapNation’s acquisition of Tarentum, demonstrating scalable growth through data-driven campaign strategy and cohort performance.
Marketing Data Scientist — Wavera / Famefy App
Social Media Tech · 2025 – Present
- Built a subscription LTV prediction system using power-law renewal modeling, EMA-based trial-conversion forecasting, and late-conversion multipliers — enabling pROI computation within days of user install via RevenueCat + AppsFlyer pipelines on BigQuery.
- Designed a multi-tier segmentation framework (country, media source, platform, product) with automated weekly cohort refresh and fallback chains, powering Looker dashboards for real-time spend optimization.
Data Analytics Consultant — P.I. Works
Telecom AI · 2018 – 2022
- Built end-to-end automated reporting pipelines — from field drive-test ingestion to CEO-level presentations — replacing manual workflows with one-click report generation across multiple cities.
- Developed radio network parameter optimization and monitoring systems with automated alarm pipelines for real-time configuration tracking.
- Delivered automated trend analysis for Beeline, Telefónica, and O2 during COVID-19, enabling data-driven network capacity planning at scale.
Consultancy
AI Consultant — Backlogix
Restaurant & Hospitality Tech · 2024 – 2025
Designed and deployed an end-to-end Text2SQL system so restaurant operators could query sales, inventory, and stock data in natural language — from schema understanding through query generation to result presentation.
AI Consultant — Pingula
EdTech · 2025
Architected an LLM pipeline for automated exam question generation from Grade 5 through university-entrance level, implementing a multi-agent system (question generator → option generator → LLM judge) grounded in official curriculum materials.
Education
M.Sc. Data Science — Sapienza University of Rome
2023 – 2025 · GPA 25.05 / 30
Thesis: 5G/6G Parameter Configuration with Agentic AI — designed RAG systems and agentic pipelines leveraging open-source and closed LLMs to automate telecom parameter configuration.
Key coursework: Statistical Methods in DS, Statistical Learning, Advanced ML, Algorithmic Data Mining, Big Data, Neural Networks, Cloud Computing.
B.Sc. Electronics & Communications Engineering — Istanbul Technical University
2014 – 2019 · GPA 3.13 / 4.00
Erasmus Exchange — VSB Technical University of Ostrava
2017 – 2018 · GPA 3.20 / 4.00
Technical skills
| Languages & tools | Python, SQL, R, Git, Docker, Bash, PostgreSQL, SQLite |
| Data science | Predictive analytics, LTV forecasting, probabilistic programming, A/B testing, MMM (LightweightMMM, Robyn), Prophet, statistical modeling, experimentation |
| AI & LLM | LangChain, LangGraph, Google ADK, Agno, RAG, Text2SQL, agentic AI, fine-tuning, NLP, prompt engineering |
| Cloud & infrastructure | GCP, AWS, BigQuery, Vertex AI, CI/CD, Streamlit, Looker |
Languages
| Language | Level |
|---|---|
| Turkish | Native |
| English | C1 — Advanced |
| Italian | B1 — Intermediate |